diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:02:58 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:02:58 +0000 |
commit | 698f8c2f01ea549d77d7dc3338a12e04c11057b9 (patch) | |
tree | 173a775858bd501c378080a10dca74132f05bc50 /vendor/bstr/scripts/regex/grapheme.sh | |
parent | Initial commit. (diff) | |
download | rustc-698f8c2f01ea549d77d7dc3338a12e04c11057b9.tar.xz rustc-698f8c2f01ea549d77d7dc3338a12e04c11057b9.zip |
Adding upstream version 1.64.0+dfsg1.upstream/1.64.0+dfsg1
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'vendor/bstr/scripts/regex/grapheme.sh')
-rw-r--r-- | vendor/bstr/scripts/regex/grapheme.sh | 50 |
1 files changed, 50 insertions, 0 deletions
diff --git a/vendor/bstr/scripts/regex/grapheme.sh b/vendor/bstr/scripts/regex/grapheme.sh new file mode 100644 index 000000000..0b2b54daa --- /dev/null +++ b/vendor/bstr/scripts/regex/grapheme.sh @@ -0,0 +1,50 @@ +#!/bin/sh + +# vim: indentexpr= nosmartindent autoindent +# vim: tabstop=2 shiftwidth=2 softtabstop=2 + +# This regex was manually written, derived from the rules in UAX #29. +# Particularly, from Table 1c, which lays out a regex for grapheme clusters. + +CR="\p{gcb=CR}" +LF="\p{gcb=LF}" +Control="\p{gcb=Control}" +Prepend="\p{gcb=Prepend}" +L="\p{gcb=L}" +V="\p{gcb=V}" +LV="\p{gcb=LV}" +LVT="\p{gcb=LVT}" +T="\p{gcb=T}" +RI="\p{gcb=RI}" +Extend="\p{gcb=Extend}" +ZWJ="\p{gcb=ZWJ}" +SpacingMark="\p{gcb=SpacingMark}" + +Any="\p{any}" +ExtendPict="\p{Extended_Pictographic}" + +echo "(?x) +$CR $LF +| +$Control +| +$Prepend* +( + ( + ($L* ($V+ | $LV $V* | $LVT) $T*) + | + $L+ + | + $T+ + ) + | + $RI $RI + | + $ExtendPict ($Extend* $ZWJ $ExtendPict)* + | + [^$Control $CR $LF] +) +[$Extend $ZWJ $SpacingMark]* +| +$Any +" |